Music Composition Based on Generative Modeling with Artificial Intelligence: Challenges, Innovations and Perspectives

Zichen Wang

doi:10.54097/e682hj48

Authors

Zichen Wang

DOI:

https://doi.org/10.54097/e682hj48

Keywords:

Artificial Intelligence, Music Generation, Generative Model, Challenges and Prospects.

Abstract

In today's era of rapid technological advancement, the emergence of artificial intelligence has brought new opportunities to the field of music creation. Music, as an important part of human culture, has expressed emotions in its own unique way since ancient times and has evolved and innovated over time. This paper introduces the progress made by artificial intelligence in music creation, describes the application and innovation of multi-granular attention Transformer, Variable Auto-Encoder (VAE), and Generative Adversarial Network (GAN) and other technological models in the field of music generation and creation, objectively analyzes the current technological bottlenecks and the dilemmas of artistic integration, and looks forward to the future development trend of the technology and the prospects of its application. The future development trend of the technology and the prospect of its application will be analyzed objectively. Through comprehensive and in-depth research, the paper will sort out the development vein of the field, provide references for subsequent related research, and promote the further integration and development of the field of music creation and artificial intelligence to meet people's growing cultural and artistic needs.

Downloads

Download data is not yet available.

References

[1] Zhang Z, Li L, Zhang J, Hu Z, Wang H, Yan C, Yang J, Qi Y, Generating High-quality Symbolic Music Using Fine-grained Discriminators, arXiv preprint, 2024.

[2] Souza G, Figueiredo F, Machado A, Guimarães D, do we need more complex representations for structure? A comparison of note duration representation for Music Transformers, arXiv preprint, 2024.

[3] Wang Y, Yang W, Dai Z, Zhang Y, Zhao K, Wang H, MeloTrans: A Text to Symbolic Music Generation Model Following Human Composition Habit, arXiv preprint, 2024.

[4] Kundu S, Singh S, Iwahori Y, Emotion-Guided Image to Music Generation, arXiv preprint, 2024.

[5] Epure E V, Meseguer Brocal G, Afchar D, Hennequin R, Harnessing High-Level Song Descriptors towards Natural Language-Based Music Recommendation, arXiv preprint, 2024.

[6] Transformer Model Explained, https://www.cnblogs.com/LXP-Never/p/15850142.html, 2024/11/28.

[7] Xiong X, Xie Z, Huang D, Zhu Y, Research on Film Music Generation Based on Multi-granularity Attention Transformer, Modern Film Technology, 2024, pp. 18 – 25.

[8] [VAE] Principles, https://www.cnblosg.com/yifanrensheng/p/13586468.html, 2024/11/28.

[9] Generative Adversarial Networks in Deep Learning in One Post | 8 Years of GANs Architecture Development, https://www/cnblogs.com/wxkang/p/17133320.html, 2024/11/28.

[10] Wang T, Jin C, Li X, Tie Y, Qi L, Multi-track Music Generation Adversarial Network Based on Transformer, Computer Applications, 2021, pp. 3585 – 3589.

[11] Wang W, Li J, Li Y, Xing, X, Generating Stylistically Adjustable Music Based on Transformer-GANs, Frontiers of Information Technology & Electronic Engineering, 2024, pp. 106 – 121.

[12] Liu Z, Liu W, Ran L, Xu K, Jiang Y, Li Q, Qiao S, An Automatic Harmony Generation Algorithm Based on Deep Reinforcement Learning, Radio Communications Technology, 2024, pp. 985 – 992.

[13] Liao Y, Yue W, Jian Y, Wang Z, Gao Y, Lu C, MICW: A Multi-instrument Music Generation Model Based on Improved Compound Words, In Proceedings of the 2022 IEEE International Conference on Multimedia & Expo Workshops (ICMEW), 2022, pp. 1 – 10.

[14] Bai Y, A Music Generation Model Based on Variational Autoencoder, Computer Knowledge and Technology, 2024, pp. 15 – 18.